Question 1

Which of the following best describes the purpose of HAC (Heteroskedasticity and Autocorrelation Consistent) standard errors in time series regression analysis?

Accepted Answer

To provide valid standard errors when the error terms are both heteroskedastic and autocorrelated.

Question 2

Given the regression model $Y_t = \beta_0 + \beta_1 X_t + u_t$, what is the formula for the OLS estimator $\hat{\beta}_1$ in large samples, as described in the study materials?

Accepted Answer

$\hat{\beta}_1 = \frac{1}{T} \sum_{t=1}^T \frac{(X_t - \bar{X})u_t}{\sigma_X^2}$

Question 3

In the context of HAC standard errors, what does the term $v_t$ represent?

Accepted Answer

$v_t = (X_t - \bar{X})u_t$

Question 4

Why are conventional heteroskedasticity-robust OLS standard errors incorrect when $u_t$ is serially correlated in time series data?

Accepted Answer

Because they do not account for the non-zero covariance between $v_t$ and $v_s$ for $t 
eq s$.

Question 5

For i.i.d. cross-sectional data, what is the value of $\rho_1$ (the correlation between $v_1$ and $v_2$) and what does this imply for the factor $f_2$?

Accepted Answer

$\rho_1 = 0$, so $f_2 = 1$

Question 6

Suppose you are analyzing quarterly sales data for a company and suspect that the error terms in your regression are autocorrelated. Which adjustment should you make to your standard error calculations?

Accepted Answer

Use HAC standard errors to account for both heteroskedasticity and autocorrelation.

Question 7

Given the formula $var(\hat{\beta}_1) = \frac{1}{T} \frac{\sigma_v^2}{(\sigma_X^2)^2} 	imes f_T$, what does the factor $f_T$ represent?

Accepted Answer

An adjustment for serial correlation in the error terms.

Question 8

Which of the following is the correct formula for $f_T$ as given in the study materials?

Accepted Answer

$f_T = 1 + 2 \sum_{j=1}^{T-1} \frac{T-j}{T} \rho_j$

Question 9

A researcher uses OLS to estimate a time series model and reports standard errors that do not account for autocorrelation. What is the likely consequence?

Accepted Answer

The reported standard errors may be biased, leading to invalid inference.

Question 10

In a time series regression with $T=2$, if $\rho_1 = 0.5$ and $\sigma_v^2 = 4$, what is $var\left(\frac{1}{T} \sum_{t=1}^T v_t\right)$?

Accepted Answer

$var\left(\frac{1}{2}(v_1 + v_2)\right) = \frac{1}{2} \sigma_v^2 + \frac{1}{2} \rho_1 \sigma_v^2 = 3$

Question 11

Which of the following statements is TRUE regarding the use of HAC standard errors in business time series analysis?

Accepted Answer

HAC standard errors are necessary when error terms exhibit both heteroskedasticity and autocorrelation.

Question 12

If $cov(v_t, v_s) 
eq 0$ for $t 
eq s$ in a time series regression, what does this imply about the error terms?

Accepted Answer

The error terms are autocorrelated.

Question 13

A business analyst is using quarterly data and finds that the OLS standard errors are much smaller than the HAC standard errors. What is the most likely explanation?

Accepted Answer

There is significant autocorrelation in the error terms.

Question 14

Which of the following is NOT a reason to use HAC standard errors in a regression analysis?

Accepted Answer

To correct for omitted variable bias.

Question 15

In the formula $var(\hat{\beta}_1) = \frac{\sigma_v^2}{T(\sigma_X^2)^2}$, which scenario is this formula valid for?

Accepted Answer

When the data are i.i.d. cross-sectional.

Question 16

A financial analyst is modeling stock returns using daily data and suspects both heteroskedasticity and autocorrelation in the residuals. What is the best approach for inference?

Accepted Answer

Use HAC standard errors for hypothesis testing.

Heteroskedasticity and Autocorrelation Consistent (HAC) Standard Errors in Regression Analysis

Study Guide - Practice Questions

Study Guide - Flashcards